Chapter 18

Free and Yet Certain

How a Free Yes Can Be Assured

The Question on the Table

If God already knows you will order the chicken, are you still free to order the fish?

It is an old puzzle. Philosophers have been wrestling with it for centuries, and the answer matters more than you might think. Because this same puzzle sits right at the heart of the hope this book defends. Over the last several chapters we have built a case. God keeps libertarian freedom (Chapter 13). He knows what every person would freely choose in any set of circumstances (Chapter 14). He uses that knowledge to craft the most persuasive possible appeal for each soul, without forcing the choice (Chapter 15). And a final, forever refusal of a God who is fully seen and fully known is so improbable as to ground real conviction (Chapter 17).

But now the sharp reader asks: wait. If each person's yes is genuinely free, how can the outcome be certain? If you can guarantee that everyone will say yes, has the freedom not been quietly removed? Does certainty not kill freedom?

I think the answer is no. And I think it is no for a precise, clear reason that has been on the table in Christian philosophy for a long time. There is a difference between something being certain and something being necessary. God can know for sure what you will freely do without making you do it. Once we see that, the whole tension dissolves. The universalist hope can be both free and assured.

Stay with me, because this chapter closes the circle. Everything in Part IV has been working toward this point. If freedom and certainty are truly compatible, then the hope we have been building can stand on both legs at once: real freedom on one side, real assurance on the other.¹

Why It Matters

The stakes are plain. If certainty and freedom are rivals, then the universalist must choose one or the other.

Choose certainty, and you land in some form of determinism. God decrees who is saved. The creature has no real say. The cross becomes a mechanism, not an invitation. This is the charge that James De Young levels when he calls universalism "more deterministic and coercive than Calvinism."² And if the shoe fits, the charge sticks.

Choose freedom, and you seem to lose the assurance. A genuine "no" is always possible. So how can you be confident that all will say yes? The best you can offer is a hope, maybe even a strong hope, but never the settled conviction that Scripture seems to promise when it says God will be "all in all" (1 Corinthians 15:28).

I believe both horns of this dilemma rest on a mistake. The mistake is treating certainty and necessity as the same thing. They are not. Once you pull them apart, you can hold freedom and assurance together without either one swallowing the other. And once you do that, the whole argument of Part IV clicks into place.

I want to be clear about what I am claiming. I am not claiming that the certainty of universal salvation has been proved beyond all question. I argued in Chapter 2 that this book's case rests on conviction, not proof. What I am claiming is narrower and, I think, more interesting: that freedom and certainty are compatible. If that is right, then the critic who says "you cannot have both" has made a logical mistake. And that matters, because once the mistake is cleared away, nothing stops the rest of the argument from doing its work.⁴

The Argument

Move 1: The Conceptual Move (Certainty Versus Necessity)

Let me start with the clearest statement I know of the point. Kenneth Keathley puts it this way:

Molinists point out that the "foreknowledge entails necessity" objection of theological fatalism confuses necessity with certainty. Certainty is a property of persons. Necessity is a property of statements. God knows all truths with certainty. In fact, God knows the contingent choices of free creatures necessarily, due to His omniscience. But God necessarily foreknowing an event does not entail or require that the event necessarily happens. Kenneth Keathley, Salvation and Sovereignty, p. 32

Read that again slowly. Keathley is drawing a clean line between two things that we often blur together. When we say "God is certain Peter will deny Him," we are saying something about God: His knowledge will not be wrong. We are not saying something about Peter: that Peter had to deny Him, that no other outcome was possible.⁵

Think of it with an everyday example. Suppose your mother knows you well enough to be certain you will eat the last piece of pie. She knows your habits, your love of pie, the look on your face when you opened the fridge. She is certain. She is right. But did she make you eat the pie? Of course not. You could have left it. You chose not to. Her certainty tracked your free choice; it did not create it.⁶

Now scale that up. God's knowledge is not like your mother's educated guess. It is perfect, infallible, and complete. But the same logical point holds. His certainty is about Him, not about the event. The event remains what it always was: the free act of the person who performed it.⁷

Timothy Stratton sharpens the same point with a technical label that is worth knowing. The confusion between certainty and necessity is what philosophers call the modal fallacy.

This kind of objection is often rooted in a misunderstanding known as the modal fallacy. Some assume that if God infallibly knows how someone will freely choose, then that person must lack the ability to choose otherwise. But this confuses certainty with necessity. God is certain that Agent P will freely choose X in circumstance C at time t, but that does not mean P is necessitated to choose X. Certainty is an epistemic property, not a causal force. Timothy Stratton, Human Freedom, Divine Knowledge, and Mere Molinism

The word modal (from the Latin modus, "mode" or "manner") simply means: about possibility and necessity. A modal fallacy is a mistake about what is possible and what is necessary. And the particular mistake here is jumping from "God knows it will happen" to "it had to happen." The jump is invalid. Certainty lives in the knower. Necessity lives in the event. They are not the same room.⁸

Move 2: The Logical Move (The Freedom-and-Foreknowledge Puzzle)

The puzzle is ancient. It traces back at least to Boethius in the sixth century, and versions of it appear in Augustine and Aquinas.⁹ In modern form it runs like this:

(1) God knows, infallibly and eternally, every choice you will ever make.

(2) If God knows you will do X, then you will do X.

(3) If you will do X, then you cannot do otherwise.

(4) If you cannot do otherwise, you are not free.

The puzzle makes it look as though foreknowledge and freedom cannot both be real. If God knows your future, your future is fixed. And if it is fixed, you are not free. This is the worry called theological fatalism.¹⁰

But the standard solution, pressed by William Lane Craig and others, targets premise (3). It contains the modal fallacy we just identified. From the fact that you will do X, it does not follow that you must do X. "Will" describes what happens. "Must" describes what has to happen. They are different claims.¹¹

Here is the key insight. God's foreknowledge depends on what you will freely do. It does not cause what you do. The arrow of explanation runs from your choice to God's knowledge, not the other way around. If you were going to choose differently, God would simply have known that instead. His knowledge adjusts to the fact, not the fact to His knowledge.¹²

Keathley illustrates this beautifully with Peter's denial. Jesus declared with certainty that Peter would deny Him three times. But, Keathley writes, "it is important to have the arrow of necessity going in the right direction. God necessarily knew what decision Simon Peter would take. It was not necessary that Simon make a particular choice. Simon could have denied Christ twice or four times (or not at all), but necessarily Jesus would have correctly known Peter's response and prophesied accordingly."¹⁴

Peter's denial was certain. God knew it would happen. Jesus even told Peter in advance. But Peter was not a puppet. He made his choice freely, out of fear and weakness. And if Peter had chosen differently, Jesus would have known that and spoken accordingly. The certainty was in God's knowledge. The freedom was in Peter's will. Both were real.¹⁵

Move 3: The Metaphysical Move (Applying This to Assurance)

Now let me take this old puzzle and apply it to the hope of this book.

In Chapter 14 we explored middle knowledge (God's knowledge of what each person would freely choose in any situation He might place them in). In Chapter 15 we explored optimal grace (God's fitting of His appeal to each person in the way most likely to win a free yes). And in Chapter 17 we argued that, when you strip away ignorance and see God face-to-face, a final and forever refusal becomes so improbable that it grounds real conviction.

Put those together. Via middle knowledge, God knows what every person would freely do under every possible arrangement of circumstances, including every possible sequence of appeals, across all the time eternity allows. He knows which arrangements lead to a free yes. And being perfectly loving and perfectly wise, He brings about the world in which those arrangements are realized.

Kirk MacGregor puts this with striking clarity in his treatment of Molinism and universal salvation:

Suppose God middle-knows that there are some feasible worlds where all are eventually freely saved, one of which God in his omnibenevolence would choose to create. If and only if this is true, God could scripturally reveal that all are eventually freely saved. This is precisely what eventual universalists insist. In my judgment, then, the only hope for eventual universalism is Molinism. Kirk MacGregor, Molinist Philosophical and Theological Ventures

I want to draw out what MacGregor is saying, because I think he is exactly right about the mechanism, even though he himself is not a universalist. (He makes clear he is an "eternalist," and I respect that.)¹⁶

Here is the logic. Among all the possible worlds God could have created, some may be "feasible worlds" where every free person eventually says yes to God. If even one such world exists in the range of what God's middle knowledge reveals as feasible, and if God is perfectly loving, then He would create it. And once He does, the outcome is certain to Him: every person will, in the end, freely come. The outcome is certain (God knows it) and yet free (each person's yes is their own, undetermined by anything outside their own will).¹⁷

Notice what has not happened. God has not overridden anyone's will. He has not forced a single choice. He has not rigged the outcome. What He has done is arrange the circumstances, the sequence of appeals, the encounters, and the time in a way that fits each person perfectly, so that each one freely chooses Him. The freedom is real. The certainty is real. And there is no contradiction between them.¹⁸

Stratton reinforces this point when he explains that on Molinism, "predestination" and "determinism" are not the same thing. To predestine is to ensure something will happen. To determine, in the philosophical sense, is to necessitate it by prior causes. God predestines the outcome. He does not determine it in the sense of causally forcing each choice. The creature still freely, indeterministically chooses. And the transition from "would freely choose" to "will freely choose" does not erase the word freely.²⁰

Move 4: The Constructive Move (Compatibility)

So here is the claim, stated as plainly as I can put it.

"God is certain that all will be saved" means: God infallibly knows the outcome. His knowledge will not be wrong. He has seen, in His middle knowledge, that the world He has created is one in which every person, across the full span of God's patient appeal, freely comes to saving faith in Christ.

"All are saved freely" means: each person's yes is their own. Nothing outside the person caused the choice. The person had the real power to refuse. They did not refuse, and God knew they would not refuse, but they could have. If they would have, God would have known that instead, and the feasibility picture would have looked different.

These two statements are compatible. They describe different things. One describes God's knowledge. The other describes the creature's will. The certainty lives in God. The freedom lives in the person. And nothing about the certainty removes the freedom, because the certainty is about the free choice, not a replacement for it.²¹

I realize this takes a moment to sit with. We are used to thinking of certainty as something that locks things in place. A certain outcome feels like a fixed outcome, and a fixed outcome feels like a forced outcome. But the feeling leads us astray. Certainty is an epistemic property, a feature of knowledge. Necessity is a metaphysical property, a feature of the event itself. They can come apart. And in God's case, they do.²²

Think of it this way. Keathley offers the example of Abraham's near-sacrifice of Isaac. The outcome was "both contingent and certain. Contingently, Abraham could have failed. But according to God's flawless foreknowledge, certainly Abraham would not."²³ The event was contingent: it depended on Abraham's will. And the event was certain: God knew infallibly how Abraham would choose. Both facts held at once. Neither canceled the other.

Now apply that same structure to the salvation of all. Each person's yes is contingent: it depends on their will. But the outcome is certain: God knows infallibly that each person will freely say it. This is not a contradiction. It is the same logic that has governed Christian thinking about foreknowledge and freedom for centuries, now applied to the widest possible scope.²⁴

Move 5: The Honest Move (The Residue)

I owe you the honest qualification.

Everything I have said in this chapter rests on a chain of premises. If God has middle knowledge (Chapter 14), and if He uses it to compose optimal grace for each person (Chapter 15), and if a final rejection of a fully seen God is vanishingly improbable (Chapter 17), then the outcome is certain to Him and free for the creature.

But each of those premises is contested. Some philosophers reject middle knowledge. Some think a creature could refuse God forever, even with full sight. The chain is strong, but it is a chain, and it is only as strong as its links.²⁵

This means the assurance I am describing is conviction-grade, not theorem-grade. It is the kind of certainty we explored in Chapter 2: the certainty that your closest friend will not betray you, not the certainty that two plus two is four. The premises add up to something overwhelming. But they do not add up to a mathematical proof. And I will not pretend they do.²⁶

This is exactly where this book parts company with David Bentley Hart. Hart argues that the rational will, fully seeing the Good, cannot finally refuse it. On his account the certainty is metaphysical: it is built into the nature of reason itself.²⁷ I admire Hart's argument, and I think he is closer to right than the traditionalist. But I think he pays too high a price. By making refusal metaphysically impossible, he removes the very freedom that makes the yes worth having. I prefer the humbler claim: refusal is so improbable, given what God knows and who God is, that conviction is well grounded, but the bare logical possibility remains. This keeps freedom real. And it keeps the hope honest.²⁸

MacGregor and Ken Perszyk both press this same point from a Molinist direction. Perszyk observes that Molinism "cannot deliver necessary universalism" but can deliver "contingent universalism (i.e., as a matter of fact all will be saved)." If contingent universalism is what God achieves, then "he would know prior to creation that all persons would (will) be saved. It would thus be something he could reveal in Scripture."²⁹ That is exactly the kind of certainty this book claims: not a theorem about the structure of reason, but a settled conviction about the choices of a God who knows everything and loves perfectly.

The Voices in the Room

Several philosophers and theologians converge on this point, each from a slightly different angle. Let me name them and say where I agree and where I part company.

William Lane Craig, though not a universalist, has done more than anyone to clarify the relationship between foreknowledge and freedom. His core argument, developed across several works, is that God's foreknowledge is not a causal constraint on the creature but a reflection of what the creature will freely do. The direction of explanation runs from the choice to the knowledge, not the other way around.³⁰ Craig himself uses middle knowledge to defend a restrictive soteriology (only some are saved), and I part from him there. But his logic is the same logic I have used in this chapter, applied to a wider scope. If God's foreknowledge of Peter's denial does not remove Peter's freedom, then God's foreknowledge of everyone's eventual free acceptance does not remove anyone's freedom either. The logic does not change when the scope widens from one soul to all souls.

Kenneth Keathley, also not a universalist, has given us one of the clearest treatments of the certainty/necessity distinction in the Molinist literature. His examples from Scripture (Peter's denial, Abraham's offering, Moses interceding for Israel, the bones of Jesus at the crucifixion) show the same pattern over and over: God knows with certainty what will happen, and the events remain genuinely contingent. The outcomes could have been different. They were not, and God knew they would not be, but the "could have been" is real and not merely verbal. Keathley calls this "the congruence of certainty with contingency," and it is the bedrock of this chapter's argument.^30b

Thomas Talbott approaches the question differently. Talbott argues that a fully informed person, seeing God without distortion, cannot finally refuse God. His case leans less on the mechanics of middle knowledge and more on the nature of love and rationality. Writing in The Inescapable Love of God, Talbott notes that middle knowledge "includes more than perfect foreknowledge; it also includes a perfect knowledge of what a person would have done freely in circumstances that do not even exist." God can base His decisions on what a person would freely do under conditions that never arise.⁴⁴ I draw on Talbott's picture of the fully informed will (Chapter 17), but I stop short of his strongest claim. I think the bare power of contrary choice remains even in the beatific vision. Talbott himself calls his view "the more recent and more consistent" Arminian picture, one on which our free choices are "genuine obstacles that God must work around as he tries to bring his loving purposes to fruition," but not obstacles that can seal an eternal fate.³¹ Where his moral case is powerful, I use it. Where it presses toward necessity, I hold back.

Thomas Flint's technical work on Molinism, especially his treatment of the relationship between God's providence and creaturely freedom, provides the deeper scaffolding for the argument. Flint shows in detail how God can exercise complete providential control over the world's outcome while leaving every creaturely choice genuinely free. The freedom is preserved because God's control works through the creature's free choices, not by overriding them. This is the philosophical backbone of what we have been calling optimal grace (Chapter 15).³²

Timothy Stratton has sharpened the certainty/necessity distinction for a new generation. His treatment of the modal fallacy, his careful separation of predestination from determinism, and his insistence that God's knowledge is an epistemic property rather than a causal force have all been important in this chapter. Stratton is not a universalist, and I am not enlisting him as one. But his tools are the right tools for the job. His framework shows that a free choice and a certain outcome are not rivals. They are partners.^32b

R. Zachary Manis, in his careful study of the problem of hell, has pressed the Molinist case even harder. Manis argues that the defender of the free will argument against universalism must, in the end, reject Molinism altogether. Why? Because once you grant that God has middle knowledge, He has the tools to ensure every person freely comes to Him. And the alternative, holding Molinism and anti-universalism together, runs into deep trouble. Either there is no feasible world where all are saved (a claim Manis finds implausible), or God chose not to create one, which compromises His goodness. Manis is not a universalist, but his conclusion is striking: Molinism and universalism sit well together, while Molinism and anti-universalism do not.⁴¹

Robin Parry reaches a similar verdict from a different starting point. Writing as an evangelical universalist, Parry observes that on Molinism, if God is omniscient, He knows which feasible world will result in all being freely saved. If He is omnipotent, He can bring it about. And if He is good, He will want to. These three premises, taken together, point in one direction. What blocks the universalist conclusion, for those who resist it, is the claim that no feasible world exists in which all freely come to Christ. But Parry finds that claim hard to swallow. He quotes Talbott's judgment that a fully informed decision to reject the gospel is so irrational as to be barely intelligible as a free choice. If even one feasible world exists where all say yes, and God is who we say He is, then He would create it. The certainty of the outcome does not override freedom. It rests on it.⁴²

Andrew Hronich, in his recent philosophical defense of the hope, develops the Molinist case with fresh detail. He walks through a sequence of situations in which God, via middle knowledge, identifies the exact circumstances under which each person would freely repent. God then brings about that set of circumstances. Given that the number of human persons is finite and the range of God's middle knowledge is infinite, Hronich finds it entirely plausible that God can compose a world in which every person freely responds to His grace.⁴³ This is the constructive picture at work: not a divine bulldozer flattening every will, but a divine weaver threading every free choice into a pattern of universal restoration.

I find myself landing closest to MacGregor's formulation. He is clear that Molinism does not by itself entail universalism. It also does not entail its denial. The question of whether all are saved is, on Molinism, "decided on scriptural grounds; philosophical considerations do not apply."³³ That strikes me as exactly right. Molinism gives the universalist the mechanism by which God could guarantee a free and universal salvation. Scripture and the character of God, explored across this book, give the reason to think He did. The mechanism and the reason come from different directions. Together they converge on one conclusion: the salvation of all is both free and assured.

Counter-Objections

This is the strongest version of the worry, and it deserves a slow answer.

The claim is that you "cannot make God wrong." And that is true. If God knows you will choose A, then you will choose A. You cannot create a world in which God believes you will choose A and you actually choose B. That would be a contradiction.

But here is where the modal fallacy creeps back in. From the fact that you cannot make God wrong, it does not follow that you cannot do otherwise. Why? Because if you were going to do otherwise, God would have known that. His belief would have been different. You are not locked in by God's belief; God's belief is locked onto you.³⁴

Let me walk through it step by step. Suppose God knows you will choose A. The question is: could you have chosen B? The answer is yes. If you would have chosen B, then God would have known you would choose B. In that scenario, God's belief about your choice would be different, because your choice would be different. Nothing prevents you from choosing B. You simply will not, and God knows that. But "will not" is not "cannot." This is precisely the distinction between certainty and necessity that we started with.³⁵

Stratton makes this point well when he notes that "some mistakenly argue that because one is not free to 'trick' God, that is, to act contrary to what God foreknows, libertarian freedom is lost. But that's a non sequitur." The inability to trick an omniscient being is not a limitation on your freedom. It is a limitation on your ability to falsify a perfect knower. Those are very different things.³⁶

This is a serious charge, and I want to answer it directly.

The difference between Molinism and Calvinist determinism is not a matter of wrapping paper. It is a matter of causal structure. On Calvinist determinism (what Stratton calls "exhaustive divine determinism"), God is the sufficient cause of every human choice. The creature's action is the necessary result of God's decree. The creature could not have done otherwise. On Molinism, God knows what the creature would freely do in each set of circumstances. God then actualizes the circumstances. But the creature's choice remains the creature's own. Nothing about the circumstances causally determines the outcome. The creature really could have chosen differently, and if she would have, God would have known that and would have actualized a different arrangement.³⁷

Stratton puts the point sharply: "Molinism affirms that God predestines all that comes to pass. But it denies that he determines all that comes to pass in the necessitarian sense. This difference is not just semantic; it's the very distinction that separates Molinism from EDD-Calvinism and gives Molinism its unique ability to affirm both divine sovereignty and human libertarian freedom."³⁸

And there is a simple test. On determinism, the creature could not have done otherwise. On Molinism, the creature could have but would not have. "Could not" and "would not" describe different things. One removes the power. The other leaves the power intact and describes how it is exercised. This book claims God achieves the outcome through the second, not the first. That is not determinism by another name. It is providence exercised through genuine freedom.³⁹

Conclusion

Let me gather the threads.

The worry that freedom and certainty are rivals rests on a confusion. Certainty is about the knower. Necessity is about the event. God can be certain about what a free creature will do without that certainty making the creature unfree. The arrow of explanation runs from the free choice to God's knowledge of it, not the reverse. God's foreknowledge tracks the choice; it does not cause it.

Applied to the universalist hope, this means: God can know with certainty that all will freely be saved, because He knows (via middle knowledge) what each person would freely choose under each possible arrangement of circumstances, and He actualizes a world in which each person, across the patient span of His appeal, freely comes to faith. The outcome is certain to God and free for the creature. Both hold at once.

This closes Part IV of the book. We have argued that libertarian freedom is worth keeping (Chapter 13), that God's middle knowledge gives Him knowledge of every "what if" (Chapter 14), that optimal grace is the non-coercive means by which He saves (Chapter 15), that the transworld-depravity objection does not defeat the hope (Chapter 16), that a forever refusal of a fully seen God is vanishingly unlikely (Chapter 17), and now that freedom and certainty can stand together (this chapter).

The case is honest about what remains. The assurance is conviction-grade, not theorem-grade. It rests on premises that are strong but not uncontested. And the bare logical possibility of a final refusal still stands, because real freedom demands it. But conviction is enough. It is what this book has claimed from the start (Chapter 2), and it is what the cumulative weight of the argument will be measured against (Chapter 29).⁴⁰

"Before a word is on my tongue, you, LORD, know it completely." (Psalm 139:4)

"I make known the end from the beginning, from ancient times, what is still to come. I say, 'My purpose will stand, and I will do all that I please.'" (Isaiah 46:10)

"All those the Father gives me will come to me, and whoever comes to me I will never drive away." (John 6:37)

I believe those three statements can all be true at once. God knows every word before it is spoken, every choice before it is made. His purpose will stand. All will come. And not one will be driven away. The knowledge does not cancel the freedom. The freedom does not cancel the certainty. Both hold, because both live in a God whose love is patient enough, and whose wisdom is wide enough, to bring every soul home without breaking a single will.

That is the hope. It is free, and it is certain. And in Parts V and VI, we will see that it has room to work (Chapters 19–21) and can answer the strongest objections the critics bring against it (Chapters 22–28).